AIDeepSeek-AI and Peking University open-source DSpark, using speculative decoding to boost LLM inference throughput by up to 400%The method also accelerates Gemma and Qwen model inference.5 Sources, 103d ago, first seen 103d agoCombined views37.2K5 Sources, first seen 103d ago407 likesCombined views37.2K5 Sources, first seen 103d ago407 likes
